Was curious what is going on so took a very fast (just a few orders on a 1 minute chart) back test regardless of settings and data quality. Here are the results :
As it is seen it opened short as well as long positions. You should check your settings (are short positions opening enabled at all in your "positions" settings)
There should be 2 different ways a trade can be entered, where the second one is optional and the user has the option to have only one buy condition (the current one), or the choice to use both. (Not sure if this is possible, making a buy condition an extern bool choice, if not I guess we need to have 2 seperate EA's)
The first buy condition (and exit condition) is the exact same as what is in code now, no change needed.
The 2nd one, is as I described in the "p.s." above, however I left out an important detail. In the code, you see the double:
double Buy1_1 = iCustom(NULL, 0, "StepMA_v7", 21, 1, 0, 0, 0, 0, True, 0, 0, Current + 1);
double Buy1_2 = iCustom(NULL, 0, "StepMA_v7", 21, 1, 0, 0, 0, 0, True, 0, 0, Current + 2);This means a trade is not entered until the Step angle has changed, comparing the priorcandle to the one 2 prior(so the current candle is not considered).
What I would like to see as an option, is these changed to 0 and +1, so the current candle is compared to the one prior. Yes, this means the condition once happens, can revert since the current candle repaints. This is why, you apply the ____pips since last candle rule ALSO, which when set properly, will avoid false signals and accomplish the goal of not entering or exiting trades late.
So to summarize and clarify, the robot opens a buy order as normal OR under the condition that the Step angle has changed on the current candle AND price has moved ____ pips since last candle where ___ is selectable.
What i have got working so far is the stops and time filter, the break even not yet working,the Ea its interesting to say the least,anyway this is a quick back test of what i got so far, and when the market opens if it is working correctly will post it.
Awesome! Liking that curve, much better than the backtest curves I had, and the forward testing results as well. (I was actually surprised as I expected the bare bones EA was going to be profitable without tweaks or filters, and looking closer I seen that delayed exits and entries was hurting the performance, as well as having no BE option or any SL in place)
Of course I will want to know which settings you used, SL, TP, any time filter etc!
I think the BE is critical for the success of the StepMA EA in particular, versus many other EA's just because of the logic. The signal is quite reliable, but many times price just doesn't go that far, before reversing. In these cases getting out with BE versus a loss (most losses are small, but sometimes they are a little bigger and overall they all add up at any rate) is going to improve the equity curve a big way.
This is step ma 7 expert, it opens and closes on change of color of step ma v7. Added a break even and step ma trailing stop, the break even is working in back testing but not so sure about the trailing stop yet, so for now considering this Ea a work in progress.Did as much back testing as was possible on my puter, but its very unreliable as usual.But what I have found is the Kv setting seems to be the key, and not so much the step ma length. Anyway please test this and let us know what you think it needs so we can develop it further. The StepMA_v7 needs to be in expert/indicators file, and would recommend to use this indicator version because did some very minor tweaks to this one.
I did something wrong, but to me the graph has only one color!
what time frame are we to test it wit ?
In the indicator external parameters change extern int ColorMode = 0; to ColorMode = 2;
Would say any time frame, but probably look for time frames with good long lasting trends to latch onto.
Great stuff thanks so much, will put it to work right away.
Can I ask what kind of tweaking was done to Indy?